Method and apparatus for predicting service life of superconducting magnet

Abstract

Disclosed in the present application are a method and apparatus for predicting the service life of a superconducting magnet. The method comprises: collecting temperature data of a superconducting magnet, and determining a first service life parameter of the superconducting magnet on the basis of the temperature data (S101); collecting magnetic field intensity data of the superconducting magnet, and determining a second service life parameter of the superconducting magnet on the basis of the magnetic field intensity data (S102); collecting air pressure data of the superconducting magnet, and determining a third service life parameter of the superconducting magnet on the basis of the air pressure data (S103); and performing prediction on the basis of the first service life parameter, the second service life parameter and the third service life parameter, so as to obtain the service life of the superconducting magnet (S104). In the present application, by means of obtaining service life prediction data of a superconducting magnet from different physical fields, a prediction result is more accurate and thus more suitable for predicting the service life of a superconducting magnet of a high-temperature superconducting electrodynamic suspension train in an offline state.

Description

A method and device for predicting the life of a superconducting magnet

[0001] Cross-references

[0002] This application claims priority to domestic application No. 202311575649.1 filed on November 23, 2023, entitled “A method and device for predicting the life of a superconducting magnet”, the entire contents of which are incorporated herein by reference. Technical Field

[0003] The present application relates to the field of data processing technology, and in particular to a method and device for predicting the life of a superconducting magnet. Background Art

[0004] High-temperature superconducting electric levitation trains require superconducting magnets to maintain a superconducting state, generating a strong electromagnetic field for levitation, traction, and guidance. Offline operation, in which the power supply is cut off, vacuum pumping is stopped, and the refrigeration unit is deactivated during train operation, is known as offline operation. While this mode significantly reduces energy consumption and improves train safety, it places higher demands on the lifespan of superconducting magnets to maintain their superconducting state.

[0005] Summary of the Invention

[0006] In response to the above problems, the present application provides a method and apparatus for predicting the life of a superconducting magnet, which improves the accuracy of predicting the life of a superconducting magnet of a high-temperature superconducting electric levitation train in an offline state.

[0007] In order to achieve the above objectives, this application provides the following technical solutions:

[0008] A method for predicting the life of a superconducting magnet, comprising:

[0009] collecting temperature data of a superconducting magnet, and determining a first life parameter of the superconducting magnet based on the temperature data;

[0010] collecting magnetic field strength data of the superconducting magnet, and determining a second life parameter of the superconducting magnet based on the magnetic field strength data;

[0011] collecting air pressure data of the superconducting magnet, and determining a third life parameter of the superconducting magnet based on the air pressure data;

[0012] The life of the superconducting magnet is predicted based on the first life parameter, the second life parameter, and the third life parameter.

[0013] Optionally, it also includes:

[0014] Extracting temperature data corresponding to the superconducting magnet from test data of the superconducting magnet and establishing a temperature data set, wherein the temperature data is data of temperature changes over time during the entire time process from the start of offline operation to the end of offline operation of the superconducting magnet;

[0015] Performing sensitivity analysis on the temperature data in the temperature data set to determine an initial training sample set of the temperature data;

[0016] Normalizing the temperature data in the initial training sample set to obtain a target training sample set;

[0017] Performing training based on the target training sample set to obtain a life prediction model based on temperature data;

[0018] Wherein, determining the first life parameter of the superconducting magnet based on the temperature data includes: processing the temperature data using the life prediction model based on temperature data to obtain the first life parameter.

[0019] Optionally, performing sensitivity analysis on the temperature data in the temperature data set to determine an initial training sample set of the temperature data includes:

[0020] Determining the influencing parameters of the temperature data at different acquisition positions at the same time in the temperature data set on the life of the superconducting magnet;

[0021] Based on the influencing parameters, a target temperature collection position is determined, and temperature data collected at the target temperature collection position is determined as an initial training sample set.

[0022] Optionally, it also includes:

[0023] Extracting magnetic field strength data corresponding to the superconducting magnet from test data of the superconducting magnet and establishing a magnetic field strength data set, wherein the magnetic field strength data is data of changes in magnetic field strength over time during the entire time process from the start of offline operation to the end of offline operation of the superconducting magnet;

[0024] performing a sensitivity analysis on the magnetic field strength data in the magnetic field strength data set to determine an initial training sample set of the magnetic field strength data;

[0025] Normalizing the magnetic field intensity data in the initial training sample set to obtain a target training sample set;

[0026] Performing training based on the target training sample set to obtain a life prediction model based on magnetic field intensity data;

[0027] The determining of the second life parameter of the superconducting magnet based on the magnetic field strength data includes: processing the magnetic field strength data using the life and vehicle model based on the magnetic field strength data to obtain the second life parameter.

[0028] Optionally, it also includes:

[0029] Extracting air pressure data corresponding to the superconducting magnet from the test data of the superconducting magnet and establishing an air pressure data set, wherein the air pressure data is data of air pressure changes over time during the entire time process from the start of offline operation to the end of offline operation of the superconducting magnet;

[0030] Performing sensitivity analysis on the air pressure data in the air pressure data set to determine an initial training sample set of the air pressure data;

[0031] Normalize the air pressure data in the initial training sample set to obtain the target training sample set;

[0032] Performing training based on the target training sample set to obtain a life prediction model based on air pressure data;

[0033] Wherein, determining the third life parameter of the superconducting magnet based on the air pressure data includes:

[0034] The air pressure data is processed using the life prediction model based on air pressure data to obtain a third life parameter.

[0035] Optionally, predicting the life of the superconducting magnet according to the first life parameter, the second life parameter, and the third life parameter includes:

[0036] respectively determining parameters of the degree of influence of the temperature data, the magnetic field strength data, and the air pressure data on the life of the superconducting magnet;

[0037] Determining a weight coefficient based on the influence degree parameter;

[0038] The first life parameter, the second life parameter, and the third life parameter are calculated using the weight coefficient to obtain the life of the superconducting magnet.

[0039] A device for predicting the life of a superconducting magnet, comprising:

[0040] a first determining unit, configured to collect temperature data of the superconducting magnet and determine a first life parameter of the superconducting magnet based on the temperature data;

[0041] a second determining unit, configured to collect magnetic field strength data of the superconducting magnet, and determine a second life parameter of the superconducting magnet based on the magnetic field strength data;

[0042] a third determining unit, configured to collect air pressure data of the superconducting magnet and determine a third life parameter of the superconducting magnet based on the air pressure data;

[0043] A prediction unit is configured to predict the life of the superconducting magnet based on the first life parameter, the second life parameter, and the third life parameter.

[0044] Optionally, the method further includes: a first model building unit, configured to:

[0045] Extracting temperature data corresponding to the superconducting magnet from test data of the superconducting magnet and establishing a temperature data set, wherein the temperature data is data of temperature changes over time during the entire time process from the start of offline operation to the end of offline operation of the superconducting magnet;

[0046] Performing sensitivity analysis on the temperature data in the temperature data set to determine an initial training sample set of the temperature data;

[0047] Normalizing the temperature data in the initial training sample set to obtain a target training sample set;

[0048] Performing training based on the target training sample set to obtain a life prediction model based on temperature data;

[0049] The first determining unit is specifically configured to process the temperature data using the life prediction model based on temperature data to obtain a first life parameter.

[0050] Optionally, the method further includes: a second model building unit, configured to:

[0051] Extracting magnetic field strength data corresponding to the superconducting magnet from test data of the superconducting magnet and establishing a magnetic field strength data set, wherein the magnetic field strength data is data of changes in magnetic field strength over time during the entire time process from the start of offline operation to the end of offline operation of the superconducting magnet;

[0052] performing a sensitivity analysis on the magnetic field strength data in the magnetic field strength data set to determine an initial training sample set of the magnetic field strength data;

[0053] Normalizing the magnetic field intensity data in the initial training sample set to obtain a target training sample set;

[0054] Performing training based on the target training sample set to obtain a life prediction model based on magnetic field intensity data;

[0055] The second determining unit is specifically configured to process the magnetic field intensity data using the life and vehicle model based on the magnetic field intensity data to obtain a second life parameter.

[0056] Optionally, the method further includes: a third model building unit, configured to:

[0057] Extracting air pressure data corresponding to the superconducting magnet from the test data of the superconducting magnet and establishing an air pressure data set, wherein the air pressure data is data of air pressure changes over time during the entire time process from the start of offline operation to the end of offline operation of the superconducting magnet;

[0058] Performing sensitivity analysis on the air pressure data in the air pressure data set to determine an initial training sample set of the air pressure data;

[0059] Normalize the air pressure data in the initial training sample set to obtain the target training sample set;

[0060] Performing training based on the target training sample set to obtain a life prediction model based on air pressure data;

[0061] The third determining unit is specifically configured to:

[0062] The air pressure data is processed using the life prediction model based on air pressure data to obtain a third life parameter.

[0063] Compared to the prior art, the present application provides a method and apparatus for predicting the lifespan of a superconducting magnet. The method comprises: collecting temperature data of the superconducting magnet and determining a first lifespan parameter of the superconducting magnet based on the temperature data; collecting magnetic field strength data of the superconducting magnet and determining a second lifespan parameter of the superconducting magnet based on the magnetic field strength data; collecting air pressure data of the superconducting magnet and determining a third lifespan parameter of the superconducting magnet based on the air pressure data; and predicting the lifespan of the superconducting magnet based on the first, second, and third lifespan parameters. By acquiring superconducting magnet lifespan prediction data from different physical fields, the present application makes the prediction results more accurate and more suitable for predicting the lifespan of superconducting magnets in an offline state for high-temperature superconducting electric levitation trains. BRIEF DESCRIPTION OF THE DRAWINGS

[0064] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without any creative work.

[0065] FIG1 is a schematic flow chart of a method for predicting the life of a superconducting magnet provided in an embodiment of the present application;

[0066] FIG2 is a schematic diagram of an application of a temperature sensor provided in an embodiment of the present application;

[0067] FIG3 is a schematic diagram of a lifespan prediction method based on a model provided in an embodiment of the present application;

[0068] FIG4 is a schematic diagram of a processing flow of a POS algorithm provided in an embodiment of the present application;

[0069] FIG5 is a schematic structural diagram of a device for predicting the life of a superconducting magnet provided in an embodiment of the present application. DETAILED DESCRIPTION

[0070] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0071] The terms "first," "second," and so on in the specification, claims, and drawings of this application are used to distinguish between different items, not to describe a specific order. Furthermore, the terms "including," "having," and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or apparatus comprising a series of steps or elements is not limited to the listed steps or elements and may include steps or elements that are not listed.

[0072] In order to facilitate the description of the embodiments of the present application, the relevant terms of the present application are now explained.

[0073] Superconducting magnet: refers to an electromagnet whose coil is made of a second-type superconductor with a high transition temperature and a particularly high critical magnetic field under ultra-deep cooling. It is also the power supply component of superconducting electric levitation trains.

[0074] Ultra-deep freezing: Generally, the working temperature is -180℃, about -200℃, and liquid nitrogen equipment is used for refrigeration.

[0075] Ultra-strong magnetic field: refers to a magnetic field of 5T (Tesla) or above generated by superconducting technology, and also includes ultra-high-strength magnetic fields generated by pulse technology, hybrid magnet technology or ultra-high power electromagnet technology.

[0076] High vacuum: When the vacuum degree is lower than 1.333×10-1~1.333×10-6Pa, it is called high vacuum.

[0077] Traction force (also called suspension force or guiding force) refers to the electromagnetic force generated by the repulsion between the superconducting magnet and the ground traction coil and figure-8 coil magnetic field.

[0078] High-Temperature Superconducting Electric Maglev Train: A new type of transportation vehicle that combines high-temperature superconducting technology with magnetic levitation technology. This technology utilizes the zero-resistance properties of high-temperature superconducting materials to allow the flow of high currents and the generation of strong magnetic fields. The interaction between the onboard superconducting magnets and the magnetic field of ground-based coils enables contactless transportation, supported, guided, and driven by magnetic forces.

[0079] Offline operation: The operating model in which the power is cut off, vacuum pumping is stopped, and the refrigeration machine is stopped during the operation of the train is the offline operation mode.

[0080] Quench: Superconductors possess three fundamental properties: zero resistance, the Meissner effect, and the Josephson effect. Critical temperature, critical magnetic field, and critical current are three key parameters of superconductors. When any of these parameters, temperature, magnetic field, or current, exceed their critical values, a superconducting magnet undergoes a phase transition, becoming a normal conductor. This process is called a quench. Maintaining a constant operating temperature, the superconductor can only stably return to the superconducting state when the current flowing through it drops below the recovery current. The magnetic energy released during the quench rapidly increases the local temperature of the magnet. Excessive temperature rise can damage the superconductor's internal structure and even burn out the magnet. Safety analysis of magnets after a quench requires understanding the quench transition process, including current decay, temperature changes, and voltage changes within the magnet. This allows for estimating the thermal and voltage shocks experienced during the quench. This allows for evaluation of magnet designs that meet both magnetic field strength and uniformity requirements, providing a basis for structural and process design.

[0081] Sensitivity analysis is a method for studying and analyzing the sensitivity of a system's (or model's) state or output changes to variations in system parameters or surrounding conditions. In optimization methods, sensitivity analysis is often used to investigate the stability of optimal solutions when raw data is inaccurate or changes occur. Sensitivity analysis can also determine which parameters have the greatest impact on a system or model.

[0082] PSO-CNN Multivariable Regression Training and Prediction Model: PSO, short for Particle Swarm Optimization, is a population-based stochastic optimization technique. PSO mimics the swarming behavior of insects, herds of animals, flocks of birds, and schools of fish. These groups collaboratively search for food, with each member continuously adapting its search pattern by learning from its own experience and that of other members. The PSO-CNN Multivariable Regression Training and Prediction Model uses the PSO particle swarm optimization convolutional neural network (CNN) algorithm to construct a regression model for data training and analysis.

[0083] The method for predicting the life of a superconducting magnet provided in an embodiment of the present application is primarily used in high-temperature superconducting electric levitation trains. The superconducting magnets of high-temperature superconducting electric levitation trains need to maintain their superconducting state in ultra-deep cold, strong magnetic fields, and high vacuum conditions. Therefore, it is necessary to obtain superconducting magnet life prediction parameters from different physical fields for more accurate life prediction. Referring to FIG1 , a flow chart of a method for predicting the life of a superconducting magnet provided in an embodiment of the present application is provided. The method may include the following steps:

[0084] S101 : Collect temperature data of a superconducting magnet, and determine a first life parameter of the superconducting magnet based on the temperature data.

[0085] The temperature data includes the temperature of the superconducting magnet coil, the temperature of the cooling structure and the temperature of the nitrogen fixing chamber. These temperature data can be collected by a temperature sensor (as shown in FIG2 ), specifically, the temperature data at the current moment can be collected. The life of the superconducting magnet is then predicted based on the collected temperature data to obtain a first life parameter, that is, the first life parameter characterizes the applicable life span of the superconducting magnet predicted by the current measured temperature data. Specifically, the first life parameter of the superconducting magnet can be predicted by a pre-established prediction model based on temperature data. The prediction model can be a model obtained by training based on existing superconducting magnet data.

[0086] S102 : Collect magnetic field strength data of the superconducting magnet, and determine a second life parameter of the superconducting magnet based on the magnetic field strength data.

[0087] Magnetic field strength data can be collected at different locations and times of the superconducting magnet using magnetic field sensors installed on the outer shell. The lifespan of the superconducting magnet is then predicted based on the collected magnetic field strength data to obtain a second lifespan parameter. This second lifespan parameter represents the applicable lifespan of the superconducting magnet predicted based on the currently measured magnetic field strength data. Specifically, the second lifespan parameter of the superconducting magnet can be predicted using a pre-established prediction model based on magnetic field strength data. This prediction model can be trained based on existing magnetic field strength data of the superconducting magnet.

[0088] S103 : Collecting air pressure data of the superconducting magnet, and determining a third life parameter of the superconducting magnet based on the air pressure data.

[0089] The air pressure data is derived from the air pressure in the evacuated area between the outer shell and the nitrogen fixation chamber. The superconducting magnet's lifespan is then predicted based on this collected air pressure data to obtain a third lifespan parameter. This third lifespan parameter represents the applicable lifespan of the superconducting magnet as predicted based on the currently measured air pressure data. Specifically, the third lifespan parameter of the superconducting magnet can be predicted using a pre-established prediction model based on air pressure data. This prediction model can be trained using existing air pressure data from superconducting magnets.

[0090] S104 : Predicting the life of the superconducting magnet based on the first life parameter, the second life parameter, and the third life parameter.

[0091] The lifespan of the superconducting magnet can be predicted based on the degree of influence of temperature data corresponding to the first lifespan parameter, magnetic field strength data corresponding to the second lifespan parameter, and air pressure data corresponding to the third lifespan parameter, respectively, such as by performing a weighted calculation based on experience to obtain the lifespan of the superconducting magnet. In one embodiment, predicting the lifespan of the superconducting magnet based on the first lifespan parameter, the second lifespan parameter, and the third lifespan parameter includes: respectively determining influence parameters of the temperature data, the magnetic field strength data, and the air pressure data on the lifespan of the superconducting magnet; determining a weight coefficient based on the influence parameters; and calculating the first lifespan parameter, the second lifespan parameter, and the third lifespan parameter using the weight coefficient to obtain the lifespan of the superconducting magnet.

[0092] For example, the first life parameter, the second life parameter and the third life parameter are expressed as L T , L M , L A , based on experience and multiplied by the weighted coefficient, the predicted life can be expressed as L = 0.5L T +0.25L M +0.25L A .

[0093] The present application provides a method for predicting the lifespan of a superconducting magnet. The method comprises: collecting temperature data of the superconducting magnet and determining a first lifespan parameter of the superconducting magnet based on the temperature data; collecting magnetic field strength data of the superconducting magnet and determining a second lifespan parameter of the superconducting magnet based on the magnetic field strength data; collecting air pressure data of the superconducting magnet and determining a third lifespan parameter of the superconducting magnet based on the air pressure data; and predicting the lifespan of the superconducting magnet based on the first, second, and third lifespan parameters. In the present application, by acquiring superconducting magnet lifespan prediction data from different physical fields, the prediction results are more accurate and more suitable for predicting the lifespan of superconducting magnets in an offline state for high-temperature superconducting electric levitation trains.

[0094] The following is a detailed description of the process of life prediction based on various physical field data. See Figure 3, which is a schematic diagram of a method for predicting the life of superconducting magnets for a high-temperature superconducting electric levitation train provided in an embodiment of the present application. In this method, the life prediction model based on the temperature data set is established to predict the life L T The lifespan L is predicted by establishing a lifespan prediction model based on the magnetic field intensity data set. M The lifespan prediction model based on the air pressure data set is used to predict the lifespan L A , after determining the weighted coefficient based on experience, the lifespan is predicted to guide the test and subsequent engineering applications. The following describes the specific process of establishing the lifespan prediction model based on different types of data sets in the embodiment of the present application.

[0095] In one embodiment, the method further includes: extracting temperature data corresponding to the superconducting magnet from the test data of the superconducting magnet and establishing a temperature data set, wherein the temperature data is data collected on the temperature change over time of the superconducting magnet from the start of offline operation to the end of offline operation; performing sensitivity analysis on the temperature data in the temperature data set to determine an initial training sample set of temperature data; normalizing the temperature data in the initial training sample set to obtain a target training sample set; and training based on the target training sample set to obtain a life prediction model based on temperature data. Determining the first life parameter of the superconducting magnet based on the temperature data includes: processing the temperature data using the life prediction model based on temperature data to obtain the first life parameter.

[0096] Furthermore, the sensitivity analysis of the temperature data in the temperature data set to determine the initial training sample set of the temperature data includes: determining the influencing parameters of the temperature data at different acquisition positions at the same time in the temperature data set on the life of the superconducting magnet; based on the influencing parameters, determining the target temperature acquisition position, and determining the temperature data acquired at the target temperature acquisition position as the initial training sample set.

[0097] For example, based on the results of preliminary superconducting magnet tests, temperature data related to superconducting magnets is extracted to establish a temperature dataset, such as Ti{t1, t2, …, tn}, i = 1, 2, …, n. A sensitivity analysis is then performed on the data in the temperature dataset to determine the most relevant temperature data as the initial training sample set. For example, the four most relevant temperature data sets can be selected, represented as Ti{t1, t2, …, tn}, i = 1, 2, …, 4. The data in the initial training sample set can be normalized using the following method:

[0098] Among them, x Rmaxand x Rmin The maximum and minimum values ​​in the temperature data are respectively, and the target training sample set is obtained after normalization, such as Ttesti{t1, t2, ..., tn}, i = 1, 2, ..., 4. The data set is randomly divided into two mutually exclusive sets, namely the training set (70%) and the validation set (30%). In the embodiment of the present application, the PSO-CNN multivariate regression training prediction model can be used to train the life prediction model for the temperature data with the highest correlation. Of course, other model structures can also be used for training, and the embodiment of the present application is not limited to this.

[0099] The PSO algorithm is an optimization algorithm developed based on the study of bird flock foraging behavior. Its fundamental concepts are derived from artificial life and evolutionary computation theory. It seeks the optimal solution through collaboration and information sharing among individuals in a flock. Before using the PSO algorithm to optimize CNN parameters, the first step is to determine the hyperparameters to be optimized. Parameter selection is crucial to model performance. Two key parameters in the CNN model are selected for optimization: batch size (λ) and number of epochs (μ). These two parameters are the targets of the PSO algorithm. The CNN hyperparameter determination process based on the PSO algorithm is shown in Figure 4.

[0100] Among them, there are five main parameters that need to be determined in the particle algorithm: inertia factor, particle feature number, maximum number of iterations, self-cognition learning factor, and group cognition learning factor. The specific settings of these parameters are shown in the following table:

[0101] Construct a CNN model with 1 input layer, 4 convolutional layers, 2 pooling layers, and 1 output layer. The output layer contains 1 neuron, which is the predicted lifespan. Use the temperature data obtained in the previous experiment to verify the accuracy of the model and predict the lifespan L. T The error is controlled within 95% and is considered qualified, otherwise iterative training is performed again.

[0102] Correspondingly, in one embodiment, the method further includes: extracting magnetic field strength data corresponding to the superconducting magnet from the test data of the superconducting magnet and establishing a magnetic field strength data set, wherein the magnetic field strength data is data collected of the change in magnetic field strength over time from the start to the end of offline operation of the superconducting magnet; performing sensitivity analysis on the magnetic field strength data in the magnetic field strength data set to determine an initial training sample set of magnetic field strength data; normalizing the magnetic field strength data in the initial training sample set to obtain a target training sample set; and training based on the target training sample set to obtain a life prediction model based on magnetic field strength data. Determining the second life parameter of the superconducting magnet based on the magnetic field strength data includes: processing the magnetic field strength data using the life and vehicle model based on the magnetic field strength data to obtain the second life parameter.

[0103] For example, based on the results of the previous test, the magnetic field strength data of the superconducting magnetic poles are extracted to establish the magnetic field strength data set Mi{m1,m2,…,mn},i=1,2. Then, through sensitivity analysis, the magnetic field data of several locations (such as 4 locations) with the greatest influence at different times are selected as the initial training sample set. The training samples are normalized and set where x Rmax and x Rmin are the maximum and minimum values ​​in the data respectively. After normalization, we get the target training sample set Mtesti{m1,m2,…,mn}, i=1,2.

[0104] The PSO-CNN regression training prediction model is used to train the life prediction model for the magnetic field strength data. The magnetic field strength data obtained in the previous test is used to verify the accuracy of the model and predict the life span L M The error is controlled within 95% and is considered qualified, otherwise iterative training is performed again.

[0105] Similarly, in one embodiment of the present application, the method further includes: extracting air pressure data corresponding to the superconducting magnet from the test data of the superconducting magnet and establishing an air pressure data set, wherein the air pressure data is the data collected of the air pressure change over time during the entire time process from the start of offline operation to the end of offline operation of the superconducting magnet; performing sensitivity analysis on the air pressure data in the air pressure data set to determine an initial training sample set of the air pressure data; normalizing the air pressure data in the initial training sample set to obtain a target training sample set; and training based on the target training sample set to obtain a life prediction model based on air pressure data. Determining the third life parameter of the superconducting magnet based on the air pressure data includes: processing the air pressure data using the life prediction model based on air pressure data to obtain the third life parameter.

[0106] For example, based on the results of the previous test, the superconducting magnet pressure data is extracted to establish the pressure data set Ai{a1, a2, ..., an}, i = 1, 2. Then the pressure data is subjected to sensitivity and normalization processing to obtain the target training sample set. In the normalization processing, let where x Rmax and x Rmin are the maximum and minimum values ​​in the data respectively. After normalization, we get the training set Atesti{a1,a2,…,an},i=1,2.

[0107] The PSO-CNN regression training prediction model is used to train the life prediction model for air pressure data. The air pressure data obtained in the previous test is used to verify the accuracy of the model and predict the life span L AThe error is controlled within 95% and is considered qualified, otherwise iterative training is performed again.

[0108] Measured temperature, magnetic field strength and air pressure data, predicted life span L T , L M , L A ; Based on experience and multiplying the weighted coefficient, the predicted lifespan is L = 0.5L T +0.25L M +0.25L A , guiding experiments and subsequent engineering applications.

[0109] When estimating the life of superconducting magnets for high-temperature superconducting electric levitation trains, traditional methods generally rely on the experience of engineers to make estimates based on the temperature rise during the test. This cannot guarantee the accuracy of the estimate, and it is impossible to predict the service life of superconducting magnets before the test. When the life prediction is too conservative, the full capabilities of the superconducting magnet cannot be utilized; when the life prediction is too aggressive, it may cause the super-large magnet to lose supersonic power during operation. Therefore, this application proposes a method for predicting the life of superconducting magnets for high-temperature superconducting electric levitation trains, which obtains superconducting magnet life prediction parameters from different physical fields to perform more accurate life prediction, filling the gap in the life prediction method for superconducting magnets of high-temperature superconducting electric levitation trains in the offline state.

[0110] In an embodiment of the present application, a device for predicting the life of a superconducting magnet is further provided, as shown in FIG5 , comprising:

[0111] A first determining unit 201 is configured to collect temperature data of a superconducting magnet and determine a first life parameter of the superconducting magnet based on the temperature data;

[0112] a second determining unit 202, configured to collect magnetic field strength data of the superconducting magnet and determine a second life parameter of the superconducting magnet based on the magnetic field strength data;

[0113] a third determining unit 203, configured to collect air pressure data of the superconducting magnet and determine a third life parameter of the superconducting magnet based on the air pressure data;

[0114] The prediction unit 204 is configured to predict the life of the superconducting magnet according to the first life parameter, the second life parameter, and the third life parameter.

[0115] Optionally, the method further includes: a first model building unit, configured to:

[0116] Extracting temperature data corresponding to the superconducting magnet from test data of the superconducting magnet and establishing a temperature data set, wherein the temperature data is data of temperature changes over time during the entire time process from the start of offline operation to the end of offline operation of the superconducting magnet;

[0117] Performing sensitivity analysis on the temperature data in the temperature data set to determine an initial training sample set of the temperature data;

[0118] Normalizing the temperature data in the initial training sample set to obtain a target training sample set;

[0119] Performing training based on the target training sample set to obtain a life prediction model based on temperature data;

[0120] The first determining unit is specifically configured to process the temperature data using the life prediction model based on temperature data to obtain a first life parameter.

[0121] Furthermore, performing sensitivity analysis on the temperature data in the temperature data set to determine an initial training sample set of the temperature data includes:

[0122] Determining the influencing parameters of the temperature data at different acquisition positions at the same time in the temperature data set on the life of the superconducting magnet;

[0123] Based on the influencing parameters, a target temperature collection position is determined, and temperature data collected at the target temperature collection position is determined as an initial training sample set.

[0124] Optionally, the method further includes: a second model building unit, configured to:

[0125] Extracting magnetic field strength data corresponding to the superconducting magnet from test data of the superconducting magnet and establishing a magnetic field strength data set, wherein the magnetic field strength data is data of changes in magnetic field strength over time during the entire time process from the start of offline operation to the end of offline operation of the superconducting magnet;

[0126] performing a sensitivity analysis on the magnetic field strength data in the magnetic field strength data set to determine an initial training sample set of the magnetic field strength data;

[0127] Normalizing the magnetic field intensity data in the initial training sample set to obtain a target training sample set;

[0128] Performing training based on the target training sample set to obtain a life prediction model based on magnetic field intensity data;

[0129] The second determining unit is specifically configured to process the magnetic field intensity data using the life and vehicle model based on the magnetic field intensity data to obtain a second life parameter.

[0130] Optionally, the method further includes: a third model building unit, configured to:

[0131] Extracting air pressure data corresponding to the superconducting magnet from the test data of the superconducting magnet and establishing an air pressure data set, wherein the air pressure data is data of air pressure changes over time during the entire time process from the start of offline operation to the end of offline operation of the superconducting magnet;

[0132] Performing sensitivity analysis on the air pressure data in the air pressure data set to determine an initial training sample set of the air pressure data;

[0133] Normalize the air pressure data in the initial training sample set to obtain the target training sample set;

[0134] Performing training based on the target training sample set to obtain a life prediction model based on air pressure data;

[0135] The third determining unit is specifically configured to:

[0136] The air pressure data is processed using the life prediction model based on air pressure data to obtain a third life parameter.

[0137] Optionally, the prediction unit is specifically configured to:

[0138] respectively determining parameters of the degree of influence of the temperature data, the magnetic field strength data, and the air pressure data on the life of the superconducting magnet;

[0139] Determining a weight coefficient based on the influence degree parameter;

[0140] The first life parameter, the second life parameter, and the third life parameter are calculated using the weight coefficient to obtain the life of the superconducting magnet.

[0141] Based on the aforementioned embodiments, embodiments of the present application provide a computer-readable storage medium storing one or more programs, which can be executed by one or more processors to implement the method for predicting the life of a superconducting magnet as described above.

[0142] An embodiment of the present application also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method for predicting the life of a superconducting magnet as described above when executing the program.

[0143] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0144] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Reference can be made to the common and similar parts between the various embodiments. For the devices disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the method description.

[0145] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present application. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application is not limited to the embodiments shown herein, but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A method for predicting the life of a superconducting magnet, characterized in that: include: collecting temperature data of the superconducting magnet, and determining a first life parameter of the superconducting magnet based on the temperature data; collecting magnetic field strength data of the superconducting magnet, and determining a second life parameter of the superconducting magnet based on the magnetic field strength data; collecting air pressure data of the superconducting magnet, and determining a third life parameter of the superconducting magnet based on the air pressure data; The life of the superconducting magnet is predicted based on the first life parameter, the second life parameter and the third life parameter.

2. The method for predicting the life of a superconducting magnet according to claim 1, characterized in that: Also includes: Extracting temperature data corresponding to the superconducting magnet from the test data of the superconducting magnet, and establishing a temperature data set, wherein the temperature data is the data of the temperature change over time of the superconducting magnet from the start of offline operation to the end of offline operation; Performing sensitivity analysis on the temperature data in the temperature data set to determine an initial training sample set of the temperature data; Normalizing the temperature data in the initial training sample set to obtain a target training sample set; Performing training based on the target training sample set to obtain a life prediction model based on temperature data; Wherein, determining the first life parameter of the superconducting magnet based on the temperature data includes: processing the temperature data using the life prediction model based on the temperature data to obtain the first life parameter.

3. The method for predicting the life of a superconducting magnet according to claim 2, characterized in that: The performing sensitivity analysis on the temperature data in the temperature data set to determine an initial training sample set of the temperature data includes: Determine whether the temperature data at different acquisition positions at the same time in the temperature data set are effective for the superconducting magnet Parameters affecting lifespan; Based on the influencing parameters, a target temperature collection position is determined, and temperature data collected at the target temperature collection position is determined as an initial training sample set.

4. The method for predicting the life of a superconducting magnet according to claim 1, characterized in that: Also includes: Extracting magnetic field strength data corresponding to the superconducting magnet from the test data of the superconducting magnet, and establishing a magnetic field strength data set, wherein the magnetic field strength data is data of the magnetic field strength change over time during the entire time process from the start of offline operation to the end of offline operation of the superconducting magnet; Performing sensitivity analysis on the magnetic field strength data in the magnetic field strength data set to determine an initial training sample set of the magnetic field strength data; Normalizing the magnetic field intensity data in the initial training sample set to obtain a target training sample set; Performing training based on the target training sample set to obtain a life prediction model based on magnetic field intensity data; Wherein, determining the second life parameter of the superconducting magnet based on the magnetic field strength data includes: processing the magnetic field strength data using the life and vehicle model based on the magnetic field strength data to obtain the second life parameter.

5. The method for predicting the life of a superconducting magnet according to claim 1, characterized in that: Also includes: Extracting air pressure data corresponding to the superconducting magnet from the test data of the superconducting magnet, and establishing an air pressure data set, wherein the air pressure data is data of air pressure variation over time during the entire time process from the start of offline operation to the end of offline operation of the superconducting magnet; Performing sensitivity analysis on the air pressure data in the air pressure data set to determine an initial training sample set of the air pressure data; Normalize the air pressure data in the initial training sample set to obtain the target training sample set; Performing training based on the target training sample set to obtain a life prediction model based on air pressure data; Wherein, determining the third life parameter of the superconducting magnet based on the air pressure data comprises: The air pressure data is processed using the life prediction model based on air pressure data to obtain a third life parameter.

6. The method for predicting the life of a superconducting magnet according to claim 1, characterized in that: The predicting of the life of the superconducting magnet according to the first life parameter, the second life parameter and the third life parameter comprises: respectively determining the parameters of the degree of influence of the temperature data, the magnetic field strength data and the air pressure data on the life of the superconducting magnet; Determining a weight coefficient based on the influence degree parameter; The first life parameter, the second life parameter and the third life parameter are calculated using the weight coefficient to obtain the life of the superconducting magnet.

7. A device for predicting the life of a superconducting magnet, characterized in that: include: a first determining unit, configured to collect temperature data of the superconducting magnet and determine a first life parameter of the superconducting magnet based on the temperature data; a second determining unit, configured to collect magnetic field strength data of the superconducting magnet, and determine a second life parameter of the superconducting magnet based on the magnetic field strength data; a third determining unit, configured to collect air pressure data of the superconducting magnet, and determine a third life parameter of the superconducting magnet based on the air pressure data; A prediction unit is used to predict the life of the superconducting magnet according to the first life parameter, the second life parameter and the third life parameter.

8. The device for predicting the life of a superconducting magnet according to claim 7, characterized in that: Also includes: The first model building unit is used to: Extracting temperature data corresponding to the superconducting magnet from the test data of the superconducting magnet, and establishing a temperature data set, wherein the temperature data is the data of the temperature change over time of the superconducting magnet from the start of offline operation to the end of offline operation; Performing sensitivity analysis on the temperature data in the temperature data set to determine an initial training sample set of the temperature data; Normalize the temperature data in the initial training sample set to obtain the target training sample set; Performing training based on the target training sample set to obtain a life prediction model based on temperature data; The first determination unit is specifically used to: process the temperature data using the life prediction model based on temperature data to obtain a first life parameter.

9. The device for predicting the life of a superconducting magnet according to claim 1, characterized in that: Also includes: The second model building unit is used for: Extracting magnetic field strength data corresponding to the superconducting magnet from the test data of the superconducting magnet, and establishing a magnetic field strength data set, wherein the magnetic field strength data is data of the magnetic field strength change over time during the entire time process from the start of offline operation to the end of offline operation of the superconducting magnet; Performing sensitivity analysis on the magnetic field strength data in the magnetic field strength data set to determine an initial training sample set of the magnetic field strength data; Normalizing the magnetic field intensity data in the initial training sample set to obtain a target training sample set; Performing training based on the target training sample set to obtain a life prediction model based on magnetic field intensity data; The second determination unit is specifically used to: process the magnetic field intensity data using the life and vehicle model based on the magnetic field intensity data to obtain a second life parameter.

10. The device for predicting the life of a superconducting magnet according to claim 7, characterized in that: Also includes: The third model building unit is used for: Extracting air pressure data corresponding to the superconducting magnet from the test data of the superconducting magnet, and establishing an air pressure data set, wherein the air pressure data is data of air pressure variation over time during the entire time process from the start of offline operation to the end of offline operation of the superconducting magnet; Performing sensitivity analysis on the air pressure data in the air pressure data set to determine an initial training sample set of the air pressure data; Normalize the air pressure data in the initial training sample set to obtain the target training sample set; Performing training based on the target training sample set to obtain a life prediction model based on air pressure data; The third determining unit is specifically used for: The air pressure data is processed using the life prediction model based on air pressure data to obtain a third life parameter.

Images